Aging a little: On the optimality of limited senescence in Escherichia coli

Abstract

Recent studies have shown that even in the absence of extrinsic stress, the morphologically symmetrically dividing model bacteria Escherichia coli do not generate offspring with equal reproductive fitness. Instead, daughter cells exhibit asymmetric division times that converge to two distinct growth states. This represents a limited senescence / rejuvenation process derived from asymmetric division that is stable for hundreds of generations. It remains unclear why the bacteria do not continue the senescence beyond this asymptote. Although there are inherent fitness benefits for heterogeneity in population growth rates, the two growth equilibria are surprisingly similar, differing by a few percent. In this work we derive an explicit model for the growth of a bacterial population with two growth equilibria, based on a generalized Fibonacci recurrence, in order to quantify the fitness benefit of a limited senescence process and examine costs associated with asymmetry that could generate the observed behavior. We find that with simple saturating effects of asymmetric partitioning of subcellular components, two distinct but similar growth states may be optimal while providing evolutionarily significant fitness advantages.

Senescence in Bacteria and Its Underlying Mechanisms

“…Considering most assumptions of these models, extreme damage segregation might not lead to maximum population growth rates, that is, perfect symmetry or complete asymmetry appears to be only favored under simplistic assumptions. Model predictions range between complete asymmetry, intermediate asymmetry, or no asymmetry (Watve et al, 2006;Ackermann et al, 2007;Evans and Steinsaltz, 2007;Erjavec et al, 2008;Chao, 2010;Rashidi et al, 2012;Clegg et al, 2014;Moger-Reischer and Lennon, 2019;Blitvić and Fernandez, 2020). This implies for populations segregating the accumulating damage in an optimal asymmetric fashion, whose lines of descendants that continuously inherit the larger fraction of damage go extinct after some variable time, while the population-consisting of many lines, including those that inherit small fractions of damage-remains viable.…”

“…Also, under such assumptions, the growth rate (fitness) of the two types of lineages is not equal since faster dividing lineages will grow faster and therefore contribute to larger fractions to the overall population. This fitness difference due to differences in generation times results in non-stable equilibria at the population level, though at the single-cell lineage level, equilibria might be observed (Blitvić and Fernandez, 2020). The model predicting two growth equilibria is supported by empirical data under benign conditions where no senescence, with respect to cell growth, has been observed (Proenca et al, 2018(Proenca et al, , 2019, though other studies performed under conditions that mirrored largely the study conditions of Proenca et al (2019) revealed senescence by illustrating declining growth rates (or increased doubling times) with age (Łapińska et al, 2019).…”

“…Such expertise in molecular biology, biophysics, computational biology, statistics, and mathematics is rarely combined in a single lab. Fortunately, integration of fluorescent markers is fairly standard for most molecular labs working on model bacteria systems (Datsenko and Wanner, 2000); the number of microfluidic setups such as the mother machine increases rapidly within and beyond biophysics labs (Wang et al, 2010;Kaiser et al, 2016;Steiner et al, 2019;Sampaio and Dunlop, 2020); the recently developed deep learning-based tools overcome the bottleneck for efficient and accurate image analysis (Lugagne et al, 2020;Ollion and Ollion, 2020); and the interest on theoretical explorations and deeper analysis of precise, high-throughput, and high-quality data remains high as evidenced by previous empirical studies that have sparked numerous theoretical models (Watve et al, 2006;Ackermann et al, 2007;Evans and Steinsaltz, 2007;Erjavec et al, 2008;Chao, 2010;Rashidi et al, 2012;Clegg et al, 2014;Robert et al, 2018;Moger-Reischer and Lennon, 2019;Steiner et al, 2019;Blitvić and Fernandez, 2020). Putting these pieces together requires joint efforts, cross talk, and collaborations that have not yet been widely established.…”

A transport approach to relate asymmetric protein segregation and population growth

“…In the next section, we apply the self-consistent Equation 10 to a toy model of E. coli recently published by Blitvic et al [13] as a simple example.…”

Statistical theory of asymmetric damage segregation in clonal cell populations

“…A number of computational and theoretical studies addressed the dynamics of asymmetric damage segregation (ADS) in a growing microbial population and their implications for the overall population fitness [7][8][9][10][11][12][13][14][15]. In [9,11] it was demonstrated numerically that asymmetry accelerates the average growth of the population as a whole.…”